Oleg Shirokikh

About Oleg Shirokikh

I build the commercial-grade platforms and orchestration layers that make complex mathematical workloads - ML, optimization, simulation, and decision models - run at scale in production. Over 13 years, my career has followed one pattern: find raw, emerging tech before it goes mainstream, architect it into permanent infrastructure, move to the next frontier. That cycle started with building a comprehensive ML suite and high-performance SDK in C++ with interfaces across C#, Java, Python, and R. Then early Big Data/Apache Spark integration when the ecosystem was still nascent. Then REST architecture, cloud infrastructure, and containerizing an entire analytical platform. Now - engineering AI orchestration layers, MCP servers, and agentic systems that control complex analytical environments and generate complete models from scratch. What holds it together: a Ph.D. in optimization and machine learning combined with deep production engineering across the full stack - from MKL-optimized linear algebra to containerized Azure runtimes to DAG flow engines to modern AI agent architectures (RAG, MCP, agentic skills). Most of these systems started the same way: I proposed the approach and built it end to end.

Experience

1. Principal Software Engineer

   Frontline Solvers

   Sep 2022 — Present · Reno, NV, US

   Proactively identified the AI opportunity and built it from prototype to full commercial product, establishing early-mover positioning in the sector.• Architecting MCP server and VS Code extensions to integrate complex analytical environments with AI agents.• Engineering generative AI pipelines that produce complete analytical models from scratch - spanning Optimization, Simulation, Decision Tables, ML, and multi-stage flows - with guided templates, example-based learning, and automated validation feedback loops.• Built a proprietary AI Agent with RAG in 2023 (OpenAI, LangChain, Python/Azure). Extended it to analyze all model types and evaluate their results.• Containerized the entire RASON REST server and client - architecting within-container solutions for database, blob storage, queues, metrics, and async model processing.• Co-invented a patent-pending system for Risk Analysis of ML models, designing and implementing the entire system end to end - bridging simulation and machine learning, including a novel synthetic data generation engine based on the Metalog distribution family.• Drove technical modernization across the team - modernized source control, led architecture and code reviews, and championed AI tooling adoption.

Education

• University of Florida

  Doctor of Philosophy (Ph.D.), Industrial and Systems Engineering

  2010 — 2013

• Voronezh State University

  Bachelor's degree, Computer Science

  2005 — 2009

• University of Florida

  Master's degree, Industrial and Systems Engineering

  2010 — 2012

• University of Florida - Warrington College of Business

  Master's Degree, Management, Finance

  2011 — 2013

Skills

• Statistics
• Machine Learning
• C#
• Information Retrieval
• Pattern Recognition
• Java
• Big Data
• Artificial Intelligence
• Apache Spark
• Data Analysis
• Mathematical Modeling
• Software Development
• Python
• Algorithm Design
• Deep Learning
• Analytics
• Matlab
• Statistical Data Analysis
• Algorithms
• Network Optimization
• Text Mining
• Graph Theory
• Neural Networks
• C++
• Predictive Modeling
• C
• Programming
• Optimization
• R
• Statistical Modeling
• Software Engineering
• Data Mining
• Scala
• Operations Research
• Sql
• Linux
• Combinatorial Optimization